To date, studies of human type 1 diabetes (T1D) have failed to provide a mechanistic understanding of the causes of the disease, largely because patients must be analyzed long after the autoimmune attack was initiated. Our ignorance of the key molecules and cells mediating the initiation and progression of human T1D may well underlie the paucity of significant new therapeutic interventions. Here we propose to reconstruct human T1D, using iPS-derived ? cells, thymic epithelial cells (TEC), and immune systems derived from T1D patients implanted in a novel immunodeficient mouse model based on the NOD-scid-IL2rgnull (NSG) strain. To accomplish our goal, we propose 2 aims. Aim 1 will validate human immune and pancreatic beta cell functions in optimized immunodeficient mice (OPTI-MICE). Currently available OPTI-MICE will be improved using genetic techniques to: 1) enhance engraftment of human cells; 2) allow for spontaneous and inducible hyperglycemia; 3) support expression of human HLA alleles and cytokines; and 4) knockout mouse genes that impair human cell engraftment and function. This suite of improvements will be validated using implantation of fetal human stem cells and tissues. Aim 2 will reconstruct human T1D in mice using cells derived from Type 1 diabetic iPS cells. These iPS cells will be used to produce the three key cell types: hematopoietic stem cells (HSC) that will generate immune systems, TEC, and ?-cells, all integral to the pathology of T1D. These cells will be derived through the use of directed differentiation and reprogramming strategies. We have been successful in generating functional human ? cells from human control and T1D patient iPS cells by a recently developed multi-step protocol for directed differentiation, providing a standardized and reproducible source of ? cells our studies. Functional human HSC will be generated using two technologies: 1) directed differentiation of iPS cells and 2) reprogramming of differentiated hematopoietic cells using defined factors to derive induced-HSCs. Functional human TEC will be generated using directed differentiation protocols similar to those used to achieve fully differentiated human ? cells. Each cell type will be subjected to rigorous analysis in vitro and in vivo to ensur full functionality. Differentiated cells derived from a single donor will be co-transplanted into OPTI-MICE, thus reconstituting an individual patient's disease in an animal model. Transplanted mice will be carefully monitored for the emergence of autoantibodies and autoreactive T-cells, and for destruction of ?-cells. This new model of human T1D will permit detailed observation, manipulation, and analysis of T1D as it progresses, enabling us to determine which cells and antigens initiate T1D. Such mechanistic insights will properly inform new approaches to curing, or even preventing, this disease. To accomplish our goal, we have assembled an interactive team of researchers formed using seed monies from the Helmsley Charitable Trust. We have now been working together and meeting regularly for over 5 years and have expertise in the relevant areas required to accomplish this project.